1.2.2 Application generation

Program code written by the programmer in high level code e.g C++

Low level code, each instruction maps to a single machine code instruction. Each machine will have its own assembly code instruction set.

Converting high level language to machine code

It is easier to learn, understand, and debug.

translates whole program into machine code into machine code in one operation, reporting errors at the end of the process.

translates and executes program into machine code line by line, stops if an error is encountered and waits for program to be debugged before running again.

Set of stages that turns source code into executable object code

Remove comments and white space. Allocate tokens. Produce a symbol table. Optimise code.

happens in lexical analysis stage when the purpose of each element is identified such as variable name, arithmetic operation, value.

Used by the complier to keep track of all the elements identified in the program.

Compiler checks that the tokens are in correct order and follow rules of the programming language being used.

Abstract syntax tree maps the structure of the program.

Used to highlight logic errors, determines whether the code that has been written has meaning within the language

Data type declaration mismatch , wrong parameters called in a subroutine, using primitive variables ( reserved words as variable names.

Seprate program which is distinct of the source code, known as object code.

Code generation. Interpreters do not produce executable file.

Happens at each stage of the compilation process, optimisers identify redundant or repeated code and remove or rearrange the code if necessary.

store of pre-written subroutines that gives access to many specialist functions for example maths linpbrary in Java.

resolves issue if several object code files when dealing with compilations of higher level languages by creating a singular executable file

Copies the executable code into the RAM,

executable code cannot be reverse engineered, source code is protected, benefits commercial developers.

faster running time, executable file doesn't need to be translated, good for performance intensive games.

program doesn’t need translation, translator doesn’t need to be installed on computer

well optimisatied

errors are reported at the end of a compilation process which means that debugging process is slower for large programs as errors can be interlinked.

recompiling to make changes

can’t be run across different processors

lexical, syntax, semantic, code generation, optimisation

line-by-line allows for interpreters allows for the developers to identify and debug errors as they occur. very useful to use in the development stage of software development.

error reports are shown in the console, debugging is quicker and easier

can be used across platforms.

source code runs more slowly that compiled code as translation process repeats itself.

program requires an interpreter.

interpreter takes up memory space.

intermediate stage used by interpreter languages to speed up the translation process.

translate assembly code to executable code

simpler process as it uses a one-to-one relationship between assembly code and machine code.